A Biosystematic Study of the African and Madagascan Rubiaceae-Anthospermeae

542 Pages·1986·25.105 MB·English

by  Dr. Christian Puff (auth.)

Plant Systematics
and Evolution
Supplementum 3
A Biosystematic Study
oft he African and
Madagascan Rubiaceae
Anthospermeae
Christian Puff
Springer-Verlag Wien New York
Dr. CHRISTIAN PUFF
Institut flir Botanik der Universitat
Wien, Austria
This work is subject to copyright.
All rights are reserved, whether the whole or part of the material is concerned, specifically
those of translation, reprinting, re-use of illustrations, broadcasting, reproduction by
photocopying machine or similar means, and storage in data banks.
© 1986 by Springer-VerlagJWien
Softcover reprint of the hardcover 1st edition 1986
With 126 Figures
Library of Congress Cataloging-in-Publication Data. Puff, Christian, 1949- . A bio
systematic study of the African and Madagascan Rubiaceae-Anthospermeae. (plant
systematics and evolution. Supplementum; 3) Bibliography: p. Includes index. 1. Rubiaceae.
2. Botany - Africa. 3. Botany - Madagascar. I. Title. II. Series.
QK495.R85P84. 1986. 583'. 52. 86-20269.
ISSN 0172-6668
ISBN -13 : 978-3-7091-8853-8 e-ISBN -13 :978-3-7091-8851-4
DOl: 10.1007/978-3-7091-8851-4
Foreword
Biosystematic studies on the Rubiaceae have a long tradition at the
Institute of Botany in Vienna. Within this family the Anthospermeae,
and especially its African and Madagascan members, are of particular
interest because of several aspects in their evolution: I) Perfection of
anemophily within an otherwise nearly exclusively zoophilous family;
2) transitions from hermaphrodity to polygamy and finally dioecy;
3) differentiation from large and long-lived shrubs to short-lived herbs;
4) adaptive radiation from humid to seasonally dry, fire-exposed and xeric
habitats. However, morphological diversity linked to sexual differentia
tion, modificatory plasticity, and eco-geographical polymorphism have
for a long time hampered our understanding of the relationships among
these African Anthospermeae.
Thus, it was imperative to put special emphasis on field observations
and to carry out a variety of experiments with cultivated plants in addition
to the analysis of an enormous herbarium material. The author, for this
reason, carried out extensive field work, often under very adverse
conditions, and covered most African countries from Ethiopia to
Southern Africa and twice visited Madagascar. In this way a multitude of
data was accumulated on the group in respect to germination and growth
form, vegetative and reproductive morphology, anatomy and biology,
embryology, karyology, crossing relationships, phytochemistry, distribu
tion and ecology, etc.
The resulting biosystematic monograph of the African Anthospermeae
covers 4 genera with 61 species. In regard to the breadth of its
multidisciplinary approach and the depth of its documentation it ranks
among the most comprehensive and advanced treatments available for
African Angiosperm groups. It is hoped that it will not only serve as a
model case for the evolutionary differentiation of such groups in tropical
Africa, but also stimulate more interest in biosystematic and monographic
studies on the flora of this continent.
Vienna, July 1986 F. EHRENDORFER
Contents
Abstract . . . . . . 1
A. Introduction. . . . 3
B. Materials and Methods 6
C. General Part. . . . 8
1. Geographical Distribution and Habitats . 8
2. Habit and Growth Form Analyses. . . 19
2.1. Large to Medium-sized Shrubs . . 19
2.2. Dwarf Shrubs. . . . . . . . 23
2.2.1. Short-lived Shrubs ("Woody Herbs") 28
2.3. Subshrubs . . . . . . . . . . . . 28
2.4. Perennial Herbs . . . . . . . . . . 29
2.5. Influence of External Factors on Habit and Growth Form. 31
2.5.1. Fire . . . . . 31
2.5.2. Browsing. . . . . . . 33
3. Stems. . . . . . . . . . . . . 34
3.1. Indumentum, Cortex, Cork . . . 34
3.2. Wood, Node and Petiole Anatomy. 34
4. Leaves . . . . . . . . . . . 37
4.1. Arrangement; Leafy Short Shoots 37
4.2. Stipules. . . . . 37
4.3. Blades and Petioles. . . . . 40
4.4. Age and Persistence. . . . . 42
4.5. Anatomy . . . . . . . . 42
4.5.1. Epidermis [incl. Trichomes and Stomata] . 42
4.5.2. Chlorenchyma. . . . . . . . . . 59
4.5.3. Bundle Sheaths and Bundle Sheath Extensions . 63
5. Inflorescences. 65
n
~~~..
n
6.1. Pedicels. .
6.2. Calyx. . . 73
6.3. Corolla . . 73
6.4. Androecium 79
6.5. Gynoecium . 80
7. Carpophores, Fruits and Seeds 81
7.1. Carpophores . . . . 81
7.2. Fruits and Seeds of AnthospermlHD, Galopina and Nenax 82
7.2.1. Indumentum . . . . . . . . 82
7.2.2. Persistent Calyx Lobes . . . . . 83
7.2.3. Fruits Dehiscing into Two Mericarps 84
7.2.4. Indehiscent Fruits. . . . 89
7.2.5. Fruit (Mericarp) Anatomy. 89
7.2.6. Seeds. . . . . . . . . . . 94
VIII Contents
7.3. Fruits and Seeds of Carpacoce 97
7.3.1. Morphology. 97
7.3.2. Anatomy. 98
7.3.3. Seeds 103
8. Embryology . . . 103
9. Karyology. . . . . 107
10. Pollen . . . . . . 122
11. Phytochemical Data. . . 125
11.1. Leaf Flavonoids (by R. D. WILSON). 125
11.2. Iridoid Glycosides . . . . . l30
12. Reproductive Biology . . . . . . . l31
12.1 Anemophily . . . . . . . . l31
12.2. Sex Distributions and Sex Ratios . l34
12.2.1. Sex Forms of Individual Plants ..... l34
12.2.2. Sex Ratios and Sex Distributions in Populations . . 141
12.2.3. Sex Distributions and Anemophily, and the Evolution
of Dioecy . . . . . . . . . . . : . . . 146
12.3. Secondary Sex Characters and Sexual Dimotphism ; . . . 148
12.4. Self-Compatibility, Auto- and Geitonogamy; Parthenocarpic
Fruits and the Question of Apomixis 152
12.5. Hybridization. . . . 156
12.6. Diaspore Dispersal . . 161
12.7. Germination Data . . 163
12.8. Regeneration after Fire. 169
D. Systematic Part. . . . . . 173
Tribe Anthospermeae, Subtribe Anthosperminae . 173
Key to the African and Madagascan Genera. 173
Anthospermum. . . . . . . . . . . . 174
Regional Keys. . . . . . . . . . . 178
A. Key to Taxa Occurring in Tropical Africa 178
B. Key to Taxa Occurring in the Flora Zambesiaca and the Flora of
Angola Area . . . . . . . . . . . . . . . . . .•. 180
C. Key to Taxa Occurring in the Flora of Southern Africa Area
Excluding the SW Cape Floristic Region. . . . . . . 181
D. Key to Taxa Occurring in the SW Cape Floristic Region 185
E. Key to Taxa Occurring in Madagascar. 189
The Taxa. . . . . . . . 190
Nenax. . . . . . . . . . 393
Key to Species and Subspecies 396
The Taxa. . . 398
Galopina. . . .. . . 427
Key to Species. . . . . . 429
The Species. . . . . . . 430
Carpacoce. . . . . . . . . 446
Key to Species and Subspecies 448
The Taxa. . . . . . . . 449
Taxa to be Excluded. . . . . 465
E. Phylogenetic Relationships and Evolutionary Aspects 467
1. Relationships Within Anthospermum . . . 467
1.1. Introductory Remarks . . . . . . 467
1.2. The Anthospermum usambarense Group. 475
Contents IX
1.3. The Anthospermum spathulatum Group. 479
1.4. The Anthospermum wbyteanum Group 481
1.5. The Anthospermum ternatum Group. . 483
1.6. The Anthospermum berbaceum Group . 485
1.7. The Anthospermum pumilum Group. . 488
1.8. Isolated SW Cape Taxa, Including General Comments on Taxa
Occurring in the SW Cape Floristic Region.. ..... 492
1.9. Isolated Madagascan Taxa. . . . .. ..... 497
1.10. Concluding Remarks . . . . . .. ..... 498
2. Affinities Between Nenax and Anthospermum, and Relationships
Within Nenax. . . . . . . . . . .. ..... 501
3. Relationships Within GaJopina and its Affinities to Other Antho
spermeae. . . . . . . . . . . . . . . . . . . . . 506
4. The Isolated Position of Carpacoce and Relationships Within the
Genus. . . . . . . . . . . . . . . . . . . 508
5. Relationships Between Anthosperminae and Extra-African Antho-
spermeae. . . . . . . . . . . . . . . 512
6. Thoughts on the Evolution of the Anthospermeae . 513
Acknowledgements 520
References . . . . . 521
Index of Taxa (in part D.) 533
Abstract
Detailed investigations of the genera Anthospermum (39 species; widely
distributed in Africa S of the Sahara and Madagascar), Nenax (11 species; S
Africa), Galopina (4 species; SE Africa) and Carpacoce (7 species; S Africa: SW
Cape Floristic Region) are presented which are based on extensive field work, the
observation of plants grown in the greenhouse and the study of fixed and dried
material in the laboratory. These investigations include growth form analyses and
notes on the influence of fire and browsing on habit and growth form; detailed
data on stem and leaf, flower, fruit and seed morphology and anatomy and surface
micromorphology; analyses of inflorescences; embryological and karyological
data and pollen measurements; phytochemical information. Numerous data on
the reproductive biology are presented: The secondary anemophily of the genera is
described in detail; information on the sex forms of individual plants, and on the
sex distributions and sex ratios within populations and species is given; the
evolution of dioecy, and the secondary sex characters and the sex dimorphism of
dioecious taxa is discussed. Evidence for the presence of self-compatibility and of
auto-and geitonogamy of taxa, gathered in greenhouse experiments, is presented;
the occurrence of parthenocarpic fruits is reported; no conclusive evidence for the
occasionally presumed presence of apomicts within Anthospermum could be
found. Experimental proof for the formation of viable hybrids between taxa, as
well as a list of putative hybrids is given. Information on the dispersal of the
diaspores, their morphological nature, and data on their germination are included.
The differential regeneration behavior of taxa after a field fire is pointed out
(plants killed by fire and regenerating from seed, vs. plants resprouting from the
base).
In the Systematic Part (D.), keys to the African and Madagascan genera of the
Anthospermeae and to the species (and subspecies) of each of the genera are given;
for the large genus Anthospermum, five "regional" keys are produced to facilitate
an easier identification of species and infraspecific taxa. Each genus is briefly
circumscribed. For each species (subspecies, variety), a detailed description is
included which takes into consideration the possible sexual dimorphism of a given
taxon and growth form differences due to external factors (fire, browsing, etc.).
Chromosome number, average pollen diameters, detailed habitat notes, a short
description of the distribution, a distribution map, a complete listing of the
specimens studied and, where appropriate, critical comments (on difficulties in the
delimitation to other taxa, on variability, on relationships, etc.) are also given.
Anthospermum has the highest concentration of taxa in the SW Cape Floristic
Region, followed by SE Africa and Madagascar (several new species endemic to
the island are described); most of the tropical taxa are confined to montane areas
and often have wide but (very) disjunct distributions.
Nenax occurs in the SW Cape Floristic Region, but several species also extend
or are confined to the drier areas to the Nand NE of that region.
Galopina is centered in SE Africa; only one, essentially afromontane species
2 Abstract
extends to the highlands of E Zimbabwe and neighbouring areas of Mozambique
and to Malawi in the N and to the Cape Region in the SW.
Carpacoce is confined to the SW Cape Floristic Region; the genus comprises
both species with a wide distribution in that area and species with very restricted or
specialized distributions.
In Part E., the evolutionary relationships within each of the genera, the close
affinities between Anthospermum and Nenax, the presumed alliances of Gaioplna,
and the isolated position of the genus Carpacoce are discussed. In the speculations
on the relationships between the genera of the Anthosperminae, the Macaronesian
genus Phyllis is also included (which, in the previous chapters, had been largely
ignored except for the occasional brief reference as it has already been studied in
detail by another author). Finally, the relationships between the Anthosperminae
and the extra-African subtribes Coprosminae and Opercuiariinae are also dealt
with, and some ideas as to the evolution and the time-space development of the
entire Anthospermeae are discussed.
A. Introduction
The family Rubiaceae is characterized by having predominantly
zoophilous flowers. Like other large families of angiosperms (the
Compositae or the Rosaceae, for example) it has, however, also produced a
number of (secondarily) anemophilous representatives. In the Rubiaceae,
it is the apparently natural tribe Anthospermeae and the mono generic tribe
Theligoneae which are comprised of exclusively wind-pollinated taxa
(PUFF 1982 a) 1.
In connection with the shift from zoophily to anemophily the genera in
question acquired a number of remarkable features which make a detailed
investigation particularly rewarding (cf. PUFF 1982 a). The tribe Antho
spermeae, widely distributed and centered in subtropical and tropical
regions of the Old and New World (cf. map, Fig. 126), consists of
approximately a dozen genera and comprises several hundred species. A
detailed study covering all of these genera would be a quite impossible task
and it was, therefore, decided to concentrate on the Anthosperminae. This
subtribe, centered in Africa and Madagascar, is comprised of the genera
Anthospermum, Nenax, Gaiopina, Carpacoce and of-as the "odd man
out" - the Macaronesian genus Phyllis. The present study deals with all
these genera except Phyllis. Frequently, however, reference to the latter, to
other extra-African Anthospermeae and sometimes also to genera of the
Paederieae, a tribe closely allied to the Anthospermeae (cf. PUFF 1982 a),
will be made for comparative reasons.
While Phyllis is well known and has been the subject of recent
investigations (MENDOZA-HEUER 1972, 1977), the remainder of the
Anthosperminae is not:
For S Africa, the most recent comprehensive treatment of the group is
that of SONDER (1865) in Flora Capensis. Some definite improvements,
primarily concerning taxa endemic to the SW Cape Floristic Region, were
published by SALTER (1937); these were also included in the Flora of the
Cape Peninsula (ADAMSON & SALTER 1950). Other than that, only few
1 Note added September 1985: The Australian monogeneric tribe Durring
tonieae (HENDERSON & GUYMER 1985) also belongs here. While there is no doubt
about the anemophily of Durringtonia, I am not fully convinced that it should be
placed into a tribe of its own. More detailed investigations would be desireable.